© Neuronpedia 2026

Privacy & Terms Blog GitHub Slack Twitter Contact

Neuronpedia

APIAssistant AxisNEW Circuit TracerNEW Steer SAE Evals Exports Community Blog Privacy & Terms Contact

LLAMA3.1-8B-IT · 23-RESID-POST-AA · 105027 ｜ Neuronpedia

Home
Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
Llama3.1-8B-IT (Instruct)
Resid Post - 131k
23-RESID-POST-AA
105027

INDEX

Explanations

Croatia

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Top Features by Cosine Similarity

Configuration

andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_23/trainer_1

Dataset (Dashboard)

Various

Features

131,072

Data Type

float32

Hook Name

blocks.23.hook_resid_post

Architecture

standard

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

Show PlotsShow ExplanationShow ActivationsShow Test FieldShow SteerShow Link

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 arsenal

-0.07

ul

-0.07

 stdin

-0.07

avis

-0.06

 Perkins

-0.06

.Cancel

-0.06

 мире

-0.06

immel

-0.06

 batteries

-0.06

um

-0.06

POSITIVE LOGITS

 Croatia

0.07

จะได

0.07

_la

0.06

dateFormat

0.06

splash

0.06

 otra

0.06

 Croatian

0.06

.OS

0.06

_WRONG

0.06

 Crosby

0.06

Activations Density 0.001%

No Known Activations